Achieving the Value of Hybrid Cloud
An open hybrid cloud platform and operating model is at the heart of a cloud strategy that returns scalable value.
The Shift to Hybrid Cloud
Companies perceive there is value if they shift the majority or all of the business technology operations to the cloud. There is clear access to innovation and agility benefits compared to traditional IT operations. As companies simplify and rationalise on-premises environments and in parallel adopt public cloud providers they can see value in addressing existing technical debt but must also recognise the danger of the rise in new technical debt.
Enterprise IT is traditionally a mix of application and infrastructure strategies, where much of their operational cost of ownership was built into the infrastructure e.g., 5x9s hardware availability and resilience. This gave rise to sourcing the necessary technology stack and lifecycle tools to maintain and operate according to the applications’ needs. With the adoption of public cloud certain SLAs often do not meet previous expectations, nor are they consistent between providers.
Cloud agility and innovation benefits come from adopting Platform based services (PaaS) that bring vertical abstraction to cloud infrastructure (IaaS). This means applications are built or sourced according to services provided by the platform. Application owners are now responsible for a continuous delivery model — the application must “keep up” as continual platform technology stack releases are provided. This is crucially important to help the enterprise maintain a healthy compliance, security and risk posture across an ever-expanding array of technologies and regulations.
Adopting hybrid cloud controls this complexity maze and delivers operational benefits beyond simply adopting PaaS. It brings horizontal simplification across all cloud environments both public cloud and on-premises. It enables commercial optionality to address concentration risk concerns and support long-term portability requirements. Hybrid cloud is successful when there is a standardisation of skills and enablement in the enterprise and its partner ecosystem, delivering IT according to a single DevSecOps led operating model.
A hybrid cloud strategy is founded on Build Once, Deploy Anywhere, and Manage Consistently.
Value Driver Model
Open hybrid cloud introduces higher degrees of standardisation in building, deploying and operating applications irrespective of their physical location (on premise, public cloud or both). Applications are deployed to a single cloud software platform. When that target is built on enterprise grade open-source software, a higher level of security currency is assured, and knowledge that its upstream open-source projects have an ongoing commitment to align to industry technology standards. You can read more on reflections on open here.
When applications are built, deployed and operated on a single software-defined platform unencumbered with location and infrastructure specific operating costs, then the overall operational cost per application lowers. This standardisation also delivers business acceleration with a step-change in release agility and clearer, defensible compliance and risk posture.
With these open hybrid cloud principles in mind, we can build a Value Driver Model to measure transformational maturity, across five dimensions.
1. Business Acceleration — focusing on release frequency and time to market of business capabilities.
2. Application Development and Maintenance (ADM) Productivity — evidenced by increasing the change to run ratio and increasing the density of applications on a standardised container platform with automated build, testing and deployment through the lifecycle environments.
3. Infrastructure Cost Efficiency — moving infrastructure deployment to be managed as code and managing that infrastructure through automation combined with lower software licensing costs and more efficient use of compute from containerisation.
4. Compliance, Regulation and Security — reduced cost of security detection and remediation through automation, unified configuration controls and configuration management. Defect density is reduced, and overall lower time and effort spent addressing security and risk controls.
5. Strategic Optionality — deliver into environments that have architectural flexibility and avoid vendor lock-in using standardised patterns for build and run, especially utilising open-source technologies that has a higher degree of evolution transparency (what will work where and in what environment); increased standardisation allowing for consistent repeatable deployment to multiple platform targets.
Achieving Value from Transformation
The value of open hybrid cloud arises from the standardisation of technology, processes and skills across all of it platforms. When teams have a single skill base and deliver to a hybrid cloud platform then the maturity levels increase.
The value achieved is a combination of benefits returned from maturing against the five levers — business acceleration; developer productivity; infrastructure cost efficiency; compliance, regulatory and security; and strategic optionality. Every organisation will have its own capacity for maturation, linked to its culture and business model, and of course different starting points and capacity for change.
Some value can absolutely be achieved by maturing application delivery that use platform services from the public cloud providers. Each public cloud provider offers different SLAs and themselves have different continual cost of operation and integrations to other platforms. This diversity and variability mean the maximum potential value will not be reached.
IBM has developed a “Hybrid Cloud Value Model”. It is an economic model based on analysis from transformations at its large-scale enterprise clients, plus industry benchmarking from external independent consultants.
It measures the maturity of the five value drivers in achieving hybrid cloud through platform service standardisation (on-premises and public cloud); a singular approach to IT delivery with DevSecOps (people, process and tools); and an enterprise-wide operating model change.
The model shows that a hybrid cloud strategy founded on a singular opinionated path can achieve 2.5x more value than utilising multiple stand-alone cloud platforms.
What this means is as you scale more application workload to the cloud AND you improve the maturity in which you do this (automation, integrated toolsets, standard way of working patterns) then a single hybrid cloud platform path will reap more benefits, compared with scaling up more and more applications on the platforms of one or more of the public clouds.
An Open Hybrid Cloud Platform
In order to take advantage of the hybrid cloud benefits at scale earlier then establishing that platform strategy early is critical. It does not mean implementation in all locations to get the benefit. Enterprises typically look for early benefit scenarios at the start of move to workload at scale under a cloud operating model e.g., consolidation of legacy workloads to a lower operational cost footprint or addressing critical applications with compliance and currency issues.
We can use a classic high level enterprise architecture overview to frame the capabilities to be developed.
Infrastructure Landing zones
You will have compute stacks that may or may not have geographic resilience built in. They require automation to build and integrate them into the single hybrid cloud platform software. The obvious starting point is to build out hybrid cloud platform onto those that are simply consumed as IaaS. Other stacks may already be offering container and other services e.g., DIY Kubernetes or public cloud services. A North Star commitment to the benefits of hybrid cloud will mean No Regrets action to convert those to a single standardised container platform service.
Standardised Container Platform
This is the heart of the strategy. There becomes a single programming model for application packaging and deployment with standard patterns of lifecycle promotion of applications into production, their container management, security etc. As the volume of workload truly starts to scale and multiple landing zones come into effect then the single management and control plane becomes important. The goal is to achieve a unified developer and operations experience and associated centre of excellence or guild to continually sustain innovation in what the platform can offer.
Consumption Strategies
Many companies have trained and practice agile delivery. With a transformation effort to move applications to a cloud operating model at scale — then addressing the skills and culture transformation is key. The strategy for consuming the open hybrid cloud platform needs to be both top-down — that operating model change — and bottom up — providing certified catalogue of software, services and DevSecOps practices by which all applications and data assets are delivered.
The Pathways to the Value of Hybrid Cloud
A basic lift and shift move of an ageing application is acceptable if necessary (e.g., burning platform issues, data center closure), as long as modernisation to containers is part of the complete migration plan. Otherwise, you are kicking the can down the road and not grasping the opportunity to simplify an application’s build and operate cost plus you are deferring the cost to bring the application compliance level up and failure cost that could haunt that uses it.
It is natural to evolve workloads that have been migrated to basic cloud hosting to their native platform container services e.g., EKS, AKS, GKE, IKS. This will result in a variety of platform software combinations for the applications across whatever clouds they are on. Teams will have different skill sets to build and operate those containerized applications. In particular you should be considering container interoperability for your multi-cloud setting. Each native platform has multiple integration points for systems management, security, especially certificate management, and container image scanning and registry and so on. This can be minimised and consolidated under a single hybrid cloud platform.
Unchecked organic evolution leads to new technical debt.
Architectural direction must be set for a cloud strategy that is looking to take advantage of the full potential of hybrid cloud benefits. It is important not to just think about adopting the Kubernetes service under your nose.
Architectural choices must consider the platform and operating model. Red Hat OpenShift presents an enterprise grade platform ecosystem — containers, security, registry, service mesh, pipelines — that are pre-integrated, built on leading open-source projects, continually tested and continually released. As a service, OpenShift can be deployed anywhere — on-premises, public cloud providers, and in edge devices.
If you combine Red Hat OpenShift with standardised DevSecOps build and deploy automation and the latest control plane lifecycle management services such as IBM Cloud Satellite, then there you can achieve a powerful unified developer and operations experience.
There are many paths in “migrating to cloud”. Once you embark on that migration at scale then the paths must converge by making architectural and technology choices that can ensure long term benefits.
This article is built of the shoulders of many that have developed and realised this approach to achieving benefits at scale in the cloud. This viewpoint is based on actual IBM experiences and trends the Cloud Engagement Hub is seeing in its clients and industry. I’d like to thank Mitch Costom, Eric Herness, Greg Hintermeister, in helping bring this viewpoint together.
